The FDA’s January 2024 guidance on the development of CAR T cell products is 34 pages of non-binding recommendations that nonetheless carry substantial operational weight. For developers navigating the space between IND submission and BLA, the guidance translates the agency’s accumulated experience reviewing six approved autologous CAR-T products into a set of expectations that, while technically optional, define the path of least resistance through regulatory review. Understanding what the guidance actually requires — and what it leaves to developer judgment — is more useful than reading it as a checklist.
Four areas deserve particular attention because they are operationally demanding, frequently underestimated, and directly connected to the clinical and commercial outcomes that make or break CAR-T programs.
Vector Copy Number: Report It Per CAR-Positive Cell, Not Per Total Cell
The guidance takes a specific position on how vector copy number should be reported: per CAR-positive cell, not per total cell in the product. This is not a trivial formatting preference. VCN per total cell and VCN per CAR-positive cell can diverge substantially when transduction efficiency is variable — which it routinely is in CAR-T manufacturing. A product with 60 percent CAR-positive cells and an apparent VCN of 1.2 per total cell actually has an average VCN of 2.0 per CAR-positive cell. That difference matters because VCN is a safety parameter: high VCN values increase the risk of insertional mutagenesis, and the risk is carried by the cells that actually express the CAR, not the total cell count.
The guidance also specifies that the VCN release criterion must be justified through a risk assessment that incorporates insertion site analysis, clonal dominance data, dose level, indication, and study population. This means VCN is not a single-threshold specification — it is a justified limit derived from a program-specific safety analysis. Developers who have treated VCN as a simple lot release number without the supporting risk assessment framework will find that BLA reviewers expect more.
The practical implication for development programs: assay development for VCN should be aligned to the CAR-positive cell fraction from the beginning. Process optimization aimed at controlling transduction efficiency — the proportion of cells that successfully incorporate the CAR construct — is a direct lever on VCN per CAR-positive cell, and the two cannot be optimized independently.
Chain of Identity: Two Independent Checks at Every Step
For autologous CAR-T products, the guidance establishes an explicit expectation for chain of identity maintenance throughout the entire manufacturing process — from leukapheresis collection through final product administration. The requirement is operationally specific: two unique identifiers on all labels, with built-in label checks incorporated into batch records prior to each processing step. At the clinical site, the same standard applies: two independent patient and label verification checks at bedside before infusion.
The chain of identity requirement reflects a real and irreversible risk. An autologous CAR-T product is manufactured from a specific patient’s cells and can only be administered to that patient. A mislabeling error at any step in the process — at collection, during manufacturing, at final product release, or at infusion — cannot be corrected after the fact. Unlike a drug product manufactured in large batches where an individual dose error is a serious but recoverable quality event, an autologous product error directly and immediately affects a specific identified patient who may have no treatment alternative.
The operational burden of this requirement extends beyond the manufacturing facility. Clinical sites administering commercial CAR-T products are required to be authorized treatment centers, and the authorization process includes verification of the site’s ability to maintain chain of identity through receipt, storage, and patient identification at infusion. Developers building commercial infrastructure need to incorporate COI verification into site training, monitoring, and quality systems — not just into the manufacturing batch record.
Fresh vs. Cryopreserved: The Guidance Takes a Side
The guidance does not treat fresh and cryopreserved CAR-T formulation as equivalent options between which developers may freely choose based on preference. It takes a position: cryopreservation is recommended when products are manufactured at a centralized location and shipped to clinical sites.
The reasoning is operationally grounded. Fresh CAR-T products have a limited shelf life — product quality degrades after formulation, and the window available for lot release testing, QC review, QA sign-off, logistics coordination, and patient preparation is compressed into that window. The guidance lists the specific elements that must be managed in that compressed window: timing for sampling and testing, QC result reporting, QA lot release review, product shipping, and clinical site receipt and handling. Each of these has its own timeline, and when any one of them encounters a delay — a QC assay requiring repeat testing, a shipping delay, a patient who is not ready for infusion on the scheduled day — the consequences for a fresh product can be irreversible.
Cryopreservation resolves most of these logistical constraints. It allows full release testing before the product leaves the manufacturing facility, provides flexibility in scheduling patient infusion, and decouples the manufacturing timeline from the clinical administration timeline. The guidance acknowledges the risks of cryopreservation — cryoprotectant toxicity, thaw process validation, cold chain management — but frames these as manageable engineering problems rather than fundamental barriers.
For developers still using fresh formulation in early clinical programs, the guidance signals that the BLA expectation is cryopreservation. Early-phase programs using fresh product should plan the transition to cryopreserved formulation and build the stability, logistics validation, and comparability data needed to support that transition before it becomes a late-stage bottleneck.
Comparability: Side-by-Side, Same Starting Material, Pre-Specified Criteria
The guidance on comparability study design for manufacturing changes is more specific than many developers anticipate. When manufacturing changes are made — to the vector, the manufacturing process, or the facility — comparability must be assessed by side-by-side testing using the same cellular starting material for both the pre-change and post-change products.
This requirement is demanding because it means the comparability study cannot be conducted retrospectively using historical lots against new lots made from different donors. The inherent donor-to-donor variability in leukapheresis starting material means that differences observed between historical and new lots cannot be attributed unambiguously to the manufacturing change versus the starting material. Side-by-side testing from the same leukapheresis collection isolates the manufacturing variable.
The guidance also specifies that comparability studies should not merely demonstrate that the post-change product meets lot release criteria. Statistical analysis with predefined acceptance criteria is required. A product that technically passes all release tests but shows a statistically significant shift in a characterization parameter — phenotype distribution, expansion kinetics, potency assay output — may still fail a properly designed comparability assessment.
The operational implication is that comparability study design should be considered at the point when manufacturing changes are being planned, not after they are implemented. The study design — which assays, which acceptance criteria, what starting material, how many lots — should be part of the change control package, not a retrospective exercise.
What the Guidance Signals About Regulatory Expectations
Taken together, these four areas — VCN reporting methodology, chain of identity, formulation strategy, and comparability study design — reflect a consistent underlying FDA position: CAR-T manufacturing complexity is not a justification for lower analytical rigor. It is the reason higher analytical rigor is required.
The programs that navigate CAR-T development most successfully are those that build the analytical and operational infrastructure to meet these expectations early — before the clinical data that makes these questions consequential is already in hand. Waiting until Phase 3 to resolve VCN reporting methodology, comparability study design, or formulation strategy means resolving them under pressure, with BLA timelines compressing the available window.
MKA Insights works with CAR-T developers at the intersection of CMC strategy, regulatory execution, and operational planning. If your program is approaching a phase transition or manufacturing change that will require comparability data or regulatory interaction, we bring the cross-functional perspective that these decisions require.